A challenge faced by commercial and retail enterprises is ticket-switching at a self-checkout. That is, a shopper may tamper the barcode on an item to his/her advantage. Part of the challenge lies in the attempt to verify the item identity claim in such scenarios. Segmenting the objects from the background is a necessary for the product verification process. However, existing segmentation systems based on background subtraction suffer from lighting changes, camera noise, and the lack of long-term background modeling, which can cause erroneous segmentations.